Brains and Machines

Phantoms in the Brain: Probing the Mysteries of the Human Mind, by V S Ramachandran and Sandra Blakeslee

2008-08-09T00:24:57+01:00

Though this book—which describes how 'freak' conditions like phantom limbs can help us understand the brain—is very interesting, it's flaws are also deep. Finally had time to finish off my review after having it bubble away in the back my brain for several months. Hope you find it useful.

If anyone knows of a much better Ramachandran that I should have read, do leave a comment. It won't make the top of my must-read list, but I'll try to have a look at it.

The iCub cometh

2008-07-17T17:15:06+01:00

I've been taking a break from writing to work on another project this spring and summer but managed to find the time to finish off a story about the iCub. This open-source robot is designed to allow academics to concentrate on implementing their theories about learning and interaction without having to focus on designing and building hardware, and is part of the general trend towards open source in the field. You can find out more by reading the full piece in EE Times.

Photo: The iCub is an artificial toddler with senses, 53 degrees of freedom, and a modular software structure designed to allow the work of different research teams to be combined.

Analog for all?

2008-02-22T17:25:51+00:00

If you've been following my analog posts you'll know that one of my concerns has been that we don't train enough engineers who are really comfortable working in this area.

Analog has two major problems: not only is it just generally more difficult to design (much harder maths!), but once you do you have to go and have a chip fabricated (i.e. spend time and money) to see how it works in practice. Digital designers have an easier job to begin with, better tools, and can reliably simulate using systems like field-programmable gate arrays (FPGAs) if they'd rather not work in software simulation alone. Plus there are a gazillion of them, which also helps them to make progress!

If you're interested in building brains into machines, this matters because analog technology seems to be the most appropriate (in terms of both power and behavior) in which to implement artifical neurons that behave in biologically-plausible or -inspired ways. This is basis of neuromorphic engineering.

Although not a new idea, field programmable analog arrays (FPAAs) may be one way of making possible both the rapid prototyping of chips and rapid training of students. Paul Hasler and his colleagues at Georgia Tech have been working on both improving the FPAA technology itself and the interfaces that designers can use with them. If you're interested in this, please check out my recent article in EE Times on the subject.

Picture: Paul Hasler and PhD student Csaba Petre demonstrate the interface to their FPAA chip. Photo by Gary Meek.

Developing systems, challenging assumptions

2008-01-11T20:26:00+00:00

I was on the phone today with Mary Lou Jepsen, founding Chief Technology Officer of One Laptop Per Child and now founder of Pixel Qi, a commercial spin-off company of OLPC that will be putting their new ambient-light-viewable displays into cell phones and laptops. Mary Lou and I both started off in holography about twenty years ago, and have (miles permitting) been friends for most of that time. And she has never ceased to amaze me: both in her talent and in her fortitude.

I haven't been following the OLPC story that closely, partly because—as a freelance—I know that it's too big a story: it's of interest to consumer journalists, tech writers, even those writing the political pages. In that kind of environment it's hard to sell freelance pieces. But talking to her today made me realize that there was a story about engineering that was worth my telling... a story about how to not just think about building a system from the ground up, but to re-think it. A story about how to innovate.

Mary Lou said that, working at OLPC, there was no single innovation that she was most proud of: it was leading the work to put the whole thing together into a practical, packaged, working system that she felt was her biggest contribution. The new machine, the XO, is the cheapest, greenest, lowest-power computer in the world. This couldn't have happened by trying to produce a stripped-down version of an expensive laptop. The whole approach had to be completely different.

That's easy to say, but I didn't really understand how deep this philosophy went until she gave me an example: she started explaining how she planned to get 5-10 times the power efficiency (and so battery life) from the new Pixel Qi laptops as from the ones we carry around today. The answer (which AMD gets and Intel doesn't!) is to turn off the central processing unit (CPU) when you don't need it.

The truth is that most of us, most of the time, have vastly more processing power than we need: like when web browsing or typing an e-mail. This is obvious. People know this. But no-one thought to question the unwritten assumption in computer design that the processor needs to be on whether it's busy or not. By finding ways to bypass the CPU when idle—by building a machine in such a way that the input/output devices can operate without it—Mary Lou and her colleages will be able to supply us with the interface we need without having to have a supercomputer running in the background for no reason.

That's an innovation. And it doesn't take too many such innovations to create something that's really new. I can't wait to see what she does next.

Photo: Mary Lou Jepsen with the XO.

Launch of The Neuromorphic Engineer

2007-11-24T23:23:16+00:00

One of the many things that have kept me from my blog in the last month or two has been working on a new progression of a newsletter I edit called The Neuromorphic Engineer. The new format is is more accessible, searchable, and generally usable, plus it allows for different kinds of content including blog posts. I've put a few of my own posts up as well as all the old newsletter archives, and there will be new content every 2-4 weeks.

If you're interested in how people are trying to build technology that emulates the neural systems of various animals (particularly, but not exclusively, using analog technology) then check this out. In general the articles are more technical than my blog, but less technical than journal papers. I'd really like to hear what you think and any suggestions you may have.